Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/283687239

Electroencephalographicneurofeedback:Levelofevidenceinmentalandbraindisordersandsuggestionsforgoodclinicalpractice

ARTICLEinNEUROPHYSIOLOGIECLINIQUE/CLINICALNEUROPHYSIOLOGY·NOVEMBER2015

ImpactFactor:1.24·DOI:10.1016/j.neucli.2015.10.077

READS

6

6AUTHORS,INCLUDING:

Jean-ArthurMicoulaudFranchi

CentreHospitalierUniversitairedeBordeaux

95PUBLICATIONS192CITATIONS

SEEPROFILE

AileenMcgonigal

AssistancePubliqueHôpitauxdeMarseille

58PUBLICATIONS1,123CITATIONS

SEEPROFILE

IlianaKotwas

FrenchNationalCentreforScientificResea…

2PUBLICATIONS4CITATIONS

SEEPROFILE

FabriceBartolomei

AssistancePubliqueHôpitauxdeMarseille

254PUBLICATIONS6,322CITATIONS

SEEPROFILE

Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,

lettingyouaccessandreadthemimmediately.

Availablefrom:Jean-ArthurMicoulaudFranchi

Retrievedon:23November2015

ARTICLE IN PRESS+ModelNEUCLI-2498; No. of Pages 11

Neurophysiologie Clinique/Clinical Neurophysiology (2015) xxx, xxx—xxx

Disponible en ligne sur

ScienceDirectwww.sciencedirect.com

REVIEW/MISE AU POINT

Electroencephalographic neurofeedback:Level of evidence in mental and braindisorders and suggestions for good clinicalpracticeNeurofeedback électroencéphalographique : niveaux depreuve clinique pour les troubles mentaux et du cerveau etpropositions de bonnes pratiques cliniques

J.-A. Micoulaud-Franchia,∗,b, A. McGonigal c,d, R. Lopeze,C. Daudet f,g, I. Kotwash, F. Bartolomeic,d,i

a Service d’explorations fonctionnelles du système nerveux, clinique du sommeil, CHU de Bordeaux, placeAmélie-Raba-Léon, 33076 Bordeaux, Franceb USR CNRS 3413 SANPSY, CHU Pellegrin, université de Bordeaux, 33076 Bordeaux, Francec Service de neurophysiologie clinique, centre hospitalo universitaire de la Timone, 264, rue Saint-Pierre,13005 Marseille, Franced Unité mixte Inserm épilepsie et cognition UMR 751, 27, boulevard Jean-Moulin, 13385 Marseille cedex 05,Francee Unité des troubles du sommeil, consultation TDA/H adulte, hôpital Gui-de-Chauliac, Montpellier, Francef Cabinet Saint-Augustin, psychiatrie, psychothérapie, Neurofeedback, 63, rue Pelouse-de-Douet, 33000Bordeaux, Franceg Université de Bordeaux, 33000 Bordeaux, Franceh Laboratoire parole et langage, UMR 7309, Aix-Marseille université, Marseille, Francei Hôpital Henri-Gastaut, établissement hospitalier spécialisé dans le traitement des épilepsies, 300,boulevard de Sainte-Marguerite, 13009 Marseille, France

Received 17 May 2015; accepted 6 October 2015

Please cite this article in press as: Micoulaud-Franchi J-A, et al. Electroencephalographic neurofeedback: Level ofevidence in mental and brain disorders and suggestions for good clinical practice. Neurophysiologie Clinique/ClinicalNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.2015.10.077

KEYWORDSNeurofeedback;Evidence-basedmedicine;

Summary The technique of electroencephalographic neurofeedback (EEG NF) emerged in the1970s and is a technique that measures a subject’s EEG signal, processes it in real time, extractsa parameter of interest and presents this information in visual or auditory form. The goal isto effectuate a behavioural modification by modulating brain activity. The EEG NF opens new

∗ Corresponding author.E-mail address: jarthur.micoulaud@gmail.com (J.-A. Micoulaud-Franchi).

http://dx.doi.org/10.1016/j.neucli.2015.10.0770987-7053/© 2015 Elsevier Masson SAS. All rights reserved.

ARTICLE IN PRESS+ModelNEUCLI-2498; No. of Pages 11

2 J.-A. Micoulaud-Franchi et al.

Model;Learning curves;Cognitiveremediation

therapeutic possibilities in the fields of psychiatry and neurology. However, the developmentof EEG NF in clinical practice requires (i) a good level of evidence of therapeutic efficacyof this technique, (ii) a good practice guide for this technique. Firstly, this article investi-gates selected trials with the following criteria: study design with controlled, randomized, andopen or blind protocol, primary endpoint related to the mental and brain disorders treatedand assessed with standardized measurement tools, identifiable EEG neurophysiological tar-gets, underpinned by pathophysiological relevance. Trials were found for: epilepsies, migraine,stroke, chronic insomnia, attentional-deficit/hyperactivity disorder (ADHD), autism spectrumdisorder, major depressive disorder, anxiety disorders, addictive disorders, psychotic disorders.Secondly, this article investigates the principles of neurofeedback therapy in line with learn-ing theory. Different underlying therapeutic models are presented didactically between twocontinua: a continuum between implicit and explicit learning and a continuum between thebiomedical model (centred on ‘‘the disease’’) and integrative biopsychosocial model of health(centred on ‘‘the illness’’). The main relevant learning model is to link neurofeedback therapywith the field of cognitive remediation techniques. The methodological specificity of neuro-feedback is to be guided by biologically relevant neurophysiological parameters. Guidelines forgood clinical practice of EEG NF concerning technical issues of electrophysiology and of learn-ing are suggested. These require validation by institutional structures for the clinical practiceof EEG NF.© 2015 Elsevier Masson SAS. All rights reserved.

MOTS CLÉSNeurofeedback ;Médecine fondée surdes preuves ;Modèle ;Courbesd’apprentissage ;Remédiationcognitive

Résumé La technique du neurofeedback électroencéphalographique (NF EEG) a émergé dansles années 1970. Il s’agit d’une technique qui mesure le signal EEG d’un sujet, le traite entemps réel, extrait un paramètre d’intérêt et présente cette information sous forme visuelleou auditive au sujet. L’objectif est de permettre une modification du comportement du sujeten apprenant à moduler l’activité de son cerveau. Le NF EEG ouvre de nouvelles possibil-ités thérapeutiques dans les domaines de la psychiatrie et de la neurologie. Cependant, ledéveloppement du NF EEG dans la pratique clinique exige : (i) un niveau satisfaisant de preuvesde l’efficacité thérapeutique de cette technique, (ii) un guide de bonnes pratiques pour cettetechnique. Premièrement, cet article a analysé les essais cliniques sélectionnés avec les critèressuivant : protocole contrôlé, randomisé, en ouvert ou en aveugle, critère principal de jugementlié aux troubles mentaux ou du cerveau évalués avec des outils de mesure validés, paramètresneurophysiologiques EEG identifiables dans la méthodologie et soutenues par une physiopatholo-gie pertinente. Des essais cliniques ont été trouvés pour : l’épilepsies, la migraine, les AVC,l’insomnie chronique, le trouble déficit de l’attention/hyperactivité (TDA/H), les troubles duspectre autistique, le trouble dépressif caractérisé, les troubles anxieux, les troubles addictifs,et les troubles psychotiques. Deuxièmement, cet article a examiné les modèles thérapeutiquesen neurofeedback en lien avec les théories de l’apprentissage. Différents modèles thérapeu-tiques sont présentés de facon didactique entre deux continuums : un continuum situé entrel’apprentissage implicite et explicite et un autre continuum situé entre un modèle biomédical(centrée sur « la maladie ») et un modèle biopsychosocial intégratif de la santé (centrée sur« le malade »). Le modèle thérapeutique le plus pertinent semble de lier les thérapies par NFEEG avec le domaine des techniques de remédiation cognitive. La spécificité méthodologiquedu NF EEG est d’être guidé par des paramètres neurophysiologiques biologiquement pertinents.Des guides de bonnes pratiques cliniques pour le NF EEG concernant les questions techniquesde l’électrophysiologie et de l’apprentissage sont proposés. La pratique clinique du NF EEGnécessite la validation de ces bonnes pratiques dans le cadre de structures institutionnellesreconnues.© 2015 Elsevier Masson SAS. Tous droits réservés.

I

Ims1[

wra1

ntroduction

t has long been recognised that brain activity may beodulated by conditioning responses. This was first demon-

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

trated in electroencephalographic (EEG) studies from the930s and 1940s with regards to the alpha-blocking response32,45,60], for example by associating an auditory stimulus

dbd

ith a light stimulus, and progressively ‘‘training’’ alphahythms to eventually attenuate to the auditory stimulationlone (see [7,6,86] for review). Subsequent studies in the960s confirmed that alpha-blocking could indeed be con-itioned [47,68] and that it was also possible to modulate

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

oth EEG synchronization and behavior using operant con-itioning; that is, increasing preferred effects and reducing

IN+Model

et

•

•

•

ip

•••

rArciir

mics

tb(t

cdIavwargcgti[ttn

E

ARTICLENEUCLI-2498; No. of Pages 11

Electroencephalographic neurofeedback

unwanted effects by using a system of reward and punish-ment [22].

While some inconsistencies exist across studies, it is nev-ertheless recognized that brain activity as measured by EEGmay thus be modified according to specific principles oflearning theory [91]. Utilising this concept of ‘‘voluntarycontrol’’ of the EEG through learning, the technique of elec-troencephalographic neurofeedback (EEG NF) emerged inthe 1970s. EEG NF is a technique that measures a subject’sEEG signal, processes it in real time, extracts a parame-ter of interest and presents this information in visual orauditory form [23]. The goal is to effectuate a behaviouralmodification by modulating brain activity [23].

The likely mechanism of this has been considered operantlearning; that is, through conditioning, the subject becomesable to increase a preferred state or behavior and decreaseundesired ones [86]. However, another viewpoint is thatthe therapeutic effect of EEG NF occurs through cognitiveremediation techniques, guided by specific and biologicallyrelevant neurophysiological parameters [86]. As such, it canbe considered within the larger spectrum of new genera-tion neurobehavioral therapies that have been developedaccording to principles of cognitive science, and that areincreasingly employed across a range of neurological, psy-chiatric and psychological conditions [87].

Following the digitization of EEG in the 1990s, therewas renewed interest in EEG NF in the context of easieraccess to electrophysiological techniques and the possibilityof more precise electrophysiological information. Increasinguse of clinical EEG NF in psychiatry and neurology since the2000s [11] has coincided with the development of NF usingfunctional magnetic resonance imaging in real time (fMRINF), which is currently used only in research, but whichoffers innovative therapeutic perspectives [48,66,98,103].Another pertinent technical advance concomitant withthe resurgence of EEG NF is the development of thebrain-computer interface (BCI), which offers original reha-bilitation and therapeutic solutions [27,111]. Despite greatinterest in fMRI NF and the BCI field of research, signifi-cant controversy exists concerning the development of EEGNF in the fields of psychiatry and neurology. With regardsto the efficacy of neurofeedback in mental and brain dis-orders, opinion within the scientific community appears tobe rather sharply divided: one school of thought is positiveand considers neurofeedback to be effective, whereas theother is very sceptical and does not consider neurofeedbacktraining to have any value, even as an adjunctive treatment.Thus, this article aims to review the evidence for (i) thera-peutic efficacy of this technique according to the principleof evidence-based medicine; (ii) good clinical practice forthis technique in line with a defined underlying therapeuticmodel and an institutional structure for regulated practice.

Level of evidence of therapeutic efficacy

There are many trials of efficacy of EEG NF in mental and

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

brain disorders, but most have significant methodologicalweaknesses [83], which could explain the scepticism of manyresearchers and clinicians concerning the effectiveness ofNF to treat mental and brain disorders. In order to rigorously

Ts[e

PRESS3

valuate the efficacy of EEG NF, in the present article onlyrials with the following criteria will be presented:

a study design with controlled, randomized, and open orblind protocol;

a primary endpoint related to the disorder treated andassessed with standardized measurement tools;

an identifiable EEG neurophysiological target, under-pinned by pathophysiological relevance (Table 1).

EEG NF neurophysiological targets can be classifiednto three categories according to three general neuro-hysiological mechanisms:

arousal [41]; emotional valence [28]; sleep.

From a simplified point of view, arousal level can beelated to the EEG power in certain spectral bands [41,75].n increase in central frontal beta band (13—30 Hz) can beelated to an increase in arousal [39] and an increase inentral frontal theta (4—8 Hz) band is related to a decreasen arousal with subjective sleepiness [102]. Moreover, anncrease in occipital alpha (8—12 Hz) can be related to aelaxed state [73].

Emotional valence has been related to EEG power asym-etry in frontal lobes [28]. Despite some inconsistencies

n the literature, an increase in alpha power in the leftompared to the right frontal cortex has been related tousceptibility toward negative emotions [97].

Sleep stages 2 and 3 (non-REM) are characterized byhe presence of sleep spindles. In animal models, it haseen found that an increase of sensory motor rhythm (SMR)Rolandic mu) can be related to an increase of spindles ando increased sleep quality [96].

Arousal is a relevant potential EEG NF target for variouslinical conditions including attention-deficit/hyperactivityisorder (ADHD), epilepsy, anxiety disorder and addiction.ndeed, ADHD symptoms can be related to a decrease inrousal [10,41]. Epilepsy can be worsened by sleep depri-ation and it has already been shown that some patientsith epilepsy can control their seizures by enhancing theirrousal level [69]. On the other hand, anxiety disorder iselated to an increase of arousal [2]. Emotional valence tar-ets are relevant for major depressive disorder, which isharacterized by emotional negative bias [29]. The SMR tar-et may be relevant for paretic after stroke, as it was foundhat SMR is related to motor imagery [72] and that motormagery could have a positive effect on stroke rehabilitation13]. Sleep targets are relevant for insomnia that is linkedo poor sleep quality [81]. (For summary see Table 1.) Thus,he particular clinical context can be used to determine theeurophysiological target selected for use with EEG NF.

pilepsies

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

he neurological disease that has benefited from the mosttudy with EEG NF is pharmacoresistant epilepsy in adults95,105]. Two meta-analyses reinforced the therapeuticffectiveness of EEG NF in pharmacoresistant epilepsy

ARTICLE IN PRESS+ModelNEUCLI-2498; No. of Pages 11

4 J.-A. Micoulaud-Franchi et al.

Table 1 EEG NF main protocols, EEG targets and supposed psychophysiological effects.

Disorder Neurophysiological targeta EEG target Aims of EEG NFtraining

Psychophysiologicaleffect of the training

ADHD andepilepsyc

Decrease of the level of reticulothalamo cortical activation(hypoarousal)b

� on fronto centralcortex

� on fronto centralcortex

� on fronto centralcortex

� on fronto centralcortex

Increase of arousal andof attentional resource

MDD Dysregulation of the balancebetween positive and negativeemotional valence (negativebias)

� on left frontalcortex

� on right frontalcortex

� on left frontalcortex

� in right frontalcortex

Inversion of the balancebetween positive andnegative emotionalvalence

GADd Cortical hyperarousal andneurovegetative hyper reactivity

�

�

� on occipitalcortex

Relaxation by decreaseof cortical hyperarousal

Insomnia Decrease of sleep spindle Rolandic mu (orSensory Motor Rhythm)

Rolandic mu Reduction of sleeplatency and increase ofsleep spindle

ADHD: attentional-deficit/hyperactivity disorder [65]; MDD: major depressive disorder [21]; GAD: generalized anxiety disorder [79,1];�: EEG spectral power in beta band > 14 Hz (associated with cortical arousal); �: EEG spectral power in theta band 3,5—7 Hz (associatedwith sleepiness); �: EEG spectral power in alpha band 8—13 Hz (associated with cortical hypometabolism).

a The identified pathophysiological target cannot be present in some patients. Its absence would be a poor prognostic factor for theEEG NF protocol.

b In ADHD and epilepsy, additional protocols were validated on slow cortical potentials (slow cortical potential, [PCS]) and wouldfocus instead on fronto-striatal networks and regulation of neuronal excitability [100]. Such a protocol was also used in migraine [90].

c The protocol used in OCD is mostly related to an EEG spectral power reduction protocol in the theta band [51].d The protocol used in addictions is close to that used in GAD but is called ‘‘deep relaxation protocol’’ because it involves increasing

7,84]

[isassttumwlaitrrgtaw

obwTpwietp

M

Ta1ttcneapepmcta

S

Tbpetf

the spectral power in the theta rather than the alpha band but [7

95,105]. However, these meta-analyses included some stud-es with poor methodological quality. Only two studies haveatisfactory methodological criteria [52,55]. A randomizednd blinded trial was performed in 8 pharmacoresistantubjects with epilepsy compared to a group of 8 matchedubjects receiving biofeedback ‘‘placebo’’ sessions. Onlyhe group with EEG NF showed a significant reduction inheir frequency of seizures (about 61%) [55]. A controlled butnrandomized trial was conducted in 41 patients with phar-acoresistant epilepsy divided into 3 groups: 34 patientsith EEG NF training, 7 patients with change of pharmaco-

ogical treatment (associated with occupational workshops),nd 11 with a placebo biofeedback training based on breath-ng techniques. Only the first two groups showed a decline inhe number of seizures (50%), which was significantly supe-ior to the placebo biofeedback [52]. This study was notandomized, but placebo scales were measured for eachroup based on questionnaires of patient satisfaction withheir treatment [52]; no difference was seen between groupsnd thus a significant placebo effect in explaining the resultsas effectively eliminated.

In line with EEG NF, a biofeedback technique focusingn skin conductance was recently studied in a randomizedlind trial, confirming the effectiveness of this technique,ith approximately 50% reduction in seizure frequency [69].his method was also effective in treating a subgroup ofatients with pharmacoresistant temporal lobe epilepsyhose seizures were triggered by stress [67]. This last study

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

s indeed of interest in its study of a specific subgroup ofpilepsy patients, since very heterogeneous populations inerms of epilepsy type and severity tend to characterizerevious EEG NF studies.

e1gO

.

igraine

reatment of migraine using biofeedback has been studied in single randomized open label study in 10 children (versus0 children on a waiting list) and was in favour of a reduc-ion in number of days with migraine per month in the groupreated with EEG NF [90]. Several other studies have beenonducted, and although in favour of efficacy, EEG NF doesot have satisfactory methodological criteria [63,99]. Thefficacy of EEG NF has yet to be confirmed, especially indults. However, it should be noted that: (i) biofeedback oneripheral parameters (skin temperature, skin conductance,lectromyography) presents a high level of evidence to sup-ort its use in migraine [71,88], especially confirmed in aeta-analysis [70]; and (ii) that neurofeedback by hemoen-

ephalography (measuring changes in blood translucencehat reflect oxygenation and local blood flow) representsn interesting avenue for the treatment of migraines [20].

troke

he use of EEG in patients with stroke was initially in reha-ilitation in the field of EEG BCI. The aim was to teach theatient to generate SMR modulation in order to control anxternal mechanical orthosis [3]. Recently, it was proposedhat SMR modulation could be also be used through EEG NFor re-education in order to improve motor function recov-

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

ry after stroke [78]. One randomized blinded trial with0 patients per group was conducted. Three interventionroups were performed: occupational therapy (OT) alone;T with EEG NF; and OT with electromyographic (EMG)

IN+Model

w[imctgp[Ncbt

A

Aconocs

M

MrNaNfTMt[

A

Wbttboibfcctsba

ARTICLENEUCLI-2498; No. of Pages 11

Electroencephalographic neurofeedback

biofeedback. EEG NF and EMG biofeedback showed similarmotor function recovery; further studies with EEG NF areneeded. It should be noted that the fields of EEG BCI andEEG NF in the context of stroke patients are closely relatedbecause of the possible recovery effect of EEG BCI througha neuroplasticity mechanism [31].

Chronic insomnia

Chronic insomnia was studied with two randomized blindedtrials with respectively 16 and 12 patients treated withEEG NF (versus 11 and 12 subjects treated with ‘‘placeboneurofeedback’’). These studies were in favour of superiorefficacy for EEG NF [42,81]. The effect of EEG NF on sleepcould also improve memory function [81].

Attentional-deficit/hyperactivity disorder

Attention-deficit/hyperactivity disorder (ADHD) in childrenis the psychiatric disorder for which EEG NF has received themost attention from society [11] and from research groups[11,35]. For example, EEG NF in ADHD received a grade1 (‘‘best support’’) rating from the American Academy ofPediatrics in 2013.

Three meta-analyses studies reinforced the therapeuticusefulness of EEG NF in ADHD [9,65,92]. The first meta-analysis of Arns et al. (2009) found an effect size (ES)that was more significant for the inattention dimensionthan for the hyperactivity dimension of ADHD [9]. However,this meta-analysis is now slightly outdated and includedsome studies with poor methodological quality. The secondmeta-analysis of Sonuga-Barke et al. (2013) found an ES sig-nificantly higher than in the control group in randomized andopen trials, but this effect remained non-significant in ran-domized and blinded trials [92]. The third meta-analysis ofMicoulaud-Franchi et al. (2014) found an ES that was signif-icantly higher than in the control group in randomized andblinded trials for the inattention dimension of ADHD [65]. Inthese two last meta-analyses, teacher assessment was cho-sen if the control group performed a cognitive remediation[34,94,104], and assessment by parents and teachers waschosen if the control group performed a placebo NF proto-col [12,54]. It should be noted that efficacy with regards tothe inattention dimension was proportional to the numberof neurofeedback sessions [9] and seemed to be maintainedover time [57].

Despite, these meta-analyses, the effectiveness of EEGNF in ADHD remain debated [8,19,26,93,107,108]. In partic-ular, the meta-analysis of Micoulaud-Franchi et al. (2014) didnot include the Arnold et al. study (2013) [4] and the stud-ies that compare EEG NF to medication [16,15,74]. Thesechoices warrant some explanations. The study by Arnoldet al. used a well-controlled, randomized and blinded pro-tocol [4]. However, the EEG NF protocol was not based onbasic learning theory (in particular by the type of reinforce-ment) used in standard EEG NF protocols [86] and the EEGrecording was carried out using an unconventional set-up

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

with electrodes placed on the forehead, a region knownto be problematic for recording because of muscle arte-facts. The Arnold et al. study thus highlights the need forgood clinical practice recommendations for EEG NF, in line

ate(

PRESS5

ith neurophysiological good practice for EEG recording30] and learning theory [86,109]. The reason for exclud-ng studies comparing EEG NF to medication is because theedication comparison group is not the most appropriate

omparison method to EEG NF. Indeed, semi-active con-rol (such as cognitive remediation techniques) and sham-NFroups to compare the EEG NF intervention are more appro-riate to control for the non-specific effects of EEG NF11]. However, comparison with medication found that EEGF was not as effective as medication in improving clini-al symptoms [16,74] or cognitive function [15], but coulde more effective on academic performance in the longererm [64].

utism spectrum disorder

utism spectrum disorder (ASD) in children is the secondondition to have benefited from a relatively large numberf efficacy studies with EEG NF [43]. However, it should beoted that, rather than demonstrating specific effectivenessn the signs of ASD, EEG NF may rather be effective on theomorbid ADHD that is present in approximately 40—50% ofubjects with ASD [43].

ajor depressive disorder

ajor depressive disorder (MDD) was studied using a singleandomized and open trial in 12 subjects treated with EEGF (versus 11 subjects treated with ‘‘placebo psychother-py’’). This study was in favour of a superior efficacy of EEGF [21]. A recent pilot study in 9 subjects reported resultsor a specific neurophysiological effect of this therapy [76].he effectiveness of EEG NF has yet to be confirmed inDD. It should be noted that fMRI NF could offer innova-

ive therapeutic perspectives in the field of mood disorders59].

nxiety disorders

ithin anxiety disorders, generalized anxiety disorder (GAD)enefited from two randomized and open trials with respec-ively nine [79] and fifteen [1] subjects in comparisono relaxation techniques with electromyographic (EMG)iofeedback in equal size groups. Obsessive-compulsive dis-rder (OCD) benefited from a randomized and blinded trialn 10 subjects (versus 10 subjects performing a biofeed-ack ‘‘placebo’’) [51]. All of these studies showed results inavour of a superior efficacy of EEG NF. However, this effi-acy should be compared to the well-established benefits ofognitive behavioral therapy (CBT). It should be noted thathe relaxation sought by EEG NF protocols used in the GADtudies is probably different from the relaxation obtainedy other techniques. EEG NF could thus be proposed as anlternative therapy for subjects who fail to achieve relax-

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

tion with other methods [85] and further research is neededo investigate the neuroplasticity effect of EEG NF in anxi-ty disorder, in particular in post-traumatic stress disorderPTSD) [49].

IN+ModelN

6

A

TptnTaa6ta[Ea

P

Pitsde[io

G

Aeptiqot[

T

Tti[

ltpbNcae

nsdN

‘caslspmisontaetnsa[

(mmia‘aEaqioviird[

pmdmflbcbrrtbttoie

ARTICLEEUCLI-2498; No. of Pages 11

ddictive disorders

he first randomized and open study with EEG NF waserformed on 15 subjects (versus 14 subjects receivingreatment as usual). All subjects were veterans of the Viet-am War, with a diagnosis of PTSD and addictive disorder.he study found a superior efficacy of EEG NF on maintainingbstinence from alcohol [77]. This effect was confirmed in

separate randomized and open trial in 61 subjects (versus0 subjects undergoing occupational support). Subjects inhis second study did not suffer from PTSD but from variousddictions (alcohol, heroin, cocaine, methamphetamine)84]. It should be noted, as for the ASD, that the effect ofEG NF in addictive disorders might be non-specific, throughn effect on anxiety disorder or ADHD comorbidities.

sychotic disorders

sychotic disorders have not so far benefited from random-zed controlled trials of EEG NF. Interestingly, it was foundhat patients with schizophrenia were capable of learningelf-control of EEG despite their cognitive and motivationaleficits [37]. An uncontrolled clinical case study suggestedfficacy of personalized EEG NF on psychotic symptoms104], and the potential therapeutic avenues for EEG NFn schizophrenia seem to be more specifically the treatmentf hallucinations using EEG or fMRI NF [62].

ood clinical practice

ny treatment must account for its model according to sci-ntific criteria of evidence and relevance and build the bestractice guide for the therapeutic technique and the insti-utional structure to support it [40]. This is particularlymportant for EEG NF. Indeed, not only the methodologicaluality of the trials is discussed in recent studies on evidencef EEG NF efficacy, but also the quality of the conduct ofhe EEG NF sessions themselves within the ‘‘active’’ group8,112].

herapeutic model

he practice of EEG NF can be examined in the light ofwo therapeutic models that can be described didactically,n terms of a spectrum between two paradigmatic poles36,101].

The first continuum is between implicit and explicitearning. Implicit learning (or ‘‘automatic’’ learning) referso the strict application of the operating conditioningaradigm that was the foundation of the first neurofeed-ack protocols. Following this therapeutic approach, EEGF sessions would target the brain in order to automaticallyorrect certain neurophysiological abnormalities. Such anpproach shares a therapeutic model close to the effect oflectrical brain stimulation [89].

Explicit (or ‘‘voluntary control’’) learning refers to cog-

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

itive remediation techniques and the development of newkills in the subject suffering from a mental and brainisorders [87]. Following this therapeutic approach, EEGF sessions would aim to teach the subject by successive

oEi(

PRESSJ.-A. Micoulaud-Franchi et al.

‘trial and error’’, then to repeat and reinforce new psy-hophysiological or neurocognitive skills (e.g. attentionalnd arousal, emotional regulation, or relaxation skills). Theubject would require a certain awareness of the skills beingearned and would then have to try to transfer these newkills into everyday life. In this model, the role of the thera-ist is essential to explain the technique, enhance learning,aintain motivation during the sessions and allow track-

ng of the performance in a given session and in successiveessions [112]. Compared to CBT techniques or other formsf cognitive remediation [87], the originality of EEG NF isot in prescribing specific cognitive tasks but in allowinghe subject to find his or her own cognitive task associ-ted with positive reinforcement and thus to maximize theffect on the neurophysiological characteristics targeted byhe neurofeedback protocol [86]. This model, which definedeurofeedback as a neurophysiological remediation, impliespecific training of the practitioner, who must continuallysk the question: ‘‘is the patient learning something?’’38,112].

The second continuum is between the biomedical modelcentred on ‘‘the disease’’) and integrative biopsychosocialodel of health (centred on ‘‘the illness’’). The biomedicalodel focuses on the brain and postulates a linear causal-

ty between a pathophysiological abnormality seen as a traitnd a mental and brain disorder. This approach would be to‘correct’’ these abnormalities in order to achieve a ther-peutic effect [36]. In this approach, a way to improveEG NF protocols would be to target ‘‘personalized’’ EEGbnormalities in a given subject, previously measured byuantitative EEG [46,106], perhaps even topographicallydentified by EEG source modelling [14,25]. Despite its the-retical appeal, this type of EEG NF protocol remains to bealidated. Moreover, the quantitative EEG technique, whichnvolves comparing the EEG data of a subject to a normat-ve EEG basis (including Z-score), raises questions about theelevance of EEG trait markers in mental and brain disor-ers [5,24], as well as questions about technical limitations18,44,46,106].

Integrative biopsychosocial models of health focus on theatient and suggest links between pathophysiological abnor-alities seen as a neurocognitive state and psychosocialimensions. Firstly, this model posits that: (i) the abnor-al neurocognitive state can be compensated by cognitiveexibility in order to ‘‘compensate’’ cognitive impairmenty ‘‘reinforcing’’ other skills [36] and (ii) the flexibilityan be enhanced by neurofeedback in order to balanceetween neuronal synchronization and desynchronization,espectively required for integration/disintegration of neu-onal assemblies during flexible cognition [80]. Secondly,his integrative model posits that this neurocognitive flexi-ility and the reinforcing of new skills would be facilitated byhe inclusion of psychosocial variables such as ‘‘locus of con-rol’’ or ‘‘self-efficacy’’ [35,86,110]. The modulating effectf these psychosocial variables on learning performance dur-ng EEG NF sessions [110], as well as on the therapeuticfficacy of the treatment, remains to be studied [35].

Currently, there are important discussions on the choice

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

f therapeutic models for defining the best practice ofEG NF. The quality of EEG NF practice does not fitnto the extremes of the spectrum. Early explicit learningwith effort) can become implicit (‘‘automatic’’) thereafter

ARTICLE IN PRESS+ModelNEUCLI-2498; No. of Pages 11

Electroencephalographic neurofeedback 7

Table 2 Signal processing software control parameters required for recording EEG in the context of EEG NF. This suggestsminimal specifications that should be available when choosing neurofeedback software.

Parameters Definitions

Filters Decomposition of the EEG signal by Fourier analysis (or other) for measuring theamplitude of the EEG spectral power in a selected frequency band

Epoch EEG epoch time in which the spectral analysis will be appliedSliding averaging Length of sliding averaging for rendering smoother amplitude fluctuations of EEG

spectral powersFeedback Feedback on the targeted EEG activity (often continuous) and positive

reinforcement (depending on the threshold and holding time)Perceptual characteristics for neurophysiological target (visual or auditorycontinuous or punctual, score or level, etc.)

Threshold Amplitude of the EEG spectral power above (or below) which positivereinforcement can be made

Occupation time Time after which the amplitude value above (or below) the threshold will lead topositive reinforcement

Inhibition Inhibition of positive reinforcement in case of change of the amplitude of thepower spectral EEG related to a biological or technical artifact, not related to aneurocognitive activity of the subject

Annotation Annotation system for a EEG-neurofeedback sessionDuration and sequence of a session Succession of training periods and rest periods and determination of their

durations and numbersSession report Evaluation of the training parameter during the session (reflection of the

performance)Storage Neurofeedback sessions with the EEG signal

Records of meetings of accounts(Note: the format of the stored files determines their exportability)

End of treatment report Evaluation of the learning curve during the sessions (evolution of the trainingparameter)

nage

•

•

•

Efadepttl(

t

Clinical data management Clinical data ma

[36,53,101] and an approach targeting a specific quantita-tive EEG brain abnormality can complete an approach takinginto account the self-efficacy of the subject. A degree oflively scientific debate provoked by different therapeuticmodels of EEG NF is in any case necessary for conscien-tious and ethical research, and practice of this therapeuticapproach [36,86,101].

Practice guideline

The rules of good practice guidelines of EEG NF concern:

• technical issues of electrophysiology [30];• technical issues of learning [86].

These aspects are rarely discussed in the EEG NF efficacytrials and there is no current consensus [11,38,61,112]. Theywill be essential to consider in order to gradually improvethe practice of EEG NF in neurology and psychiatry [40].The number of sessions of neurofeedback is usually 20—30sessions, one to three times per week, but the ideal num-ber and the optimum spacing have not yet been defined

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

[101].From the electrophysiological point of view, the practice

of neurofeedback requires high quality recording of EEG sig-nal [30]. The practitioner must:

k

•

ment system coupled with neurofeedback sessions

be familiar with the recording equipment and the entireacquisition chain (amplifier, filters, high pass and low passdigital analogue converter);

correctly set up the recording environment, especiallyconcerning placement of EEG electrodes, to obtain cor-rect impedances, and EEG montages to obtain validdeviation;

be able to read and continuously monitor the EEG to iden-tify artifacts and adjust the set-up of the material.

As part of the explicit model of learning, the practice ofEG NF requires the measurement of a parameter derivedrom the EEG and its representation through understand-ble information for the subject, in order to promote theevelopment of new neurocognitive skills transferable toveryday life. Each software package marketed for theractice of EEG NF has different advantages and disadvan-ages regarding the handling of the control parameters. Theherapist needs to manipulate these parameters to enhanceearning and motivation of the subject during the sessionsTable 2).

Good knowledge of electrophysiological and learningechniques will be particularly useful with regards to four

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

ey elements of neurofeedback sessions:

information: this is necessary before (explanation), during(motivation) and after (explanation) each session in order

IN+ModelN

8

•

•

•

I

Taniilr

tjcui

•

•

C

Irstrceioes

Naaonnaictntatticiop

D

T

R

ARTICLEEUCLI-2498; No. of Pages 11

for the subject to understand the principle of electrophys-iological measurement and to adhere to the therapeuticmodel involved in the course. However, it seems advisableto minimize the information about the type of neurocog-nitive strategies prior to a session [50,101];

the choice of threshold: this is an essential step. Adjustinga threshold (and a given occupation time) determines thenumber of positive reinforcements required to strengthenthe subject in a type of neurocognitive strategy. Thethreshold may be set automatically or manually. The man-ual threshold seems to lead to better learning [86,101]but, unlike the automatic threshold, requires performinga baseline measurement before each session;

the type of positive reinforcement: this can be visual orauditory, proportioned (graduated) or binary (present orabsent), immediate or delayed, simple or complex, fre-quent or rare. Visual feedback, which is proportionate,immediate and simple, seems to allow better learning[101]. The number of reinforcements must be sufficientto maintain the motivation of the subject but not so greatthat they alter learning [101,38]. Note that positive rein-forcement embedded in an entertaining interface (such asvideo games) may increase the motivation of the subjectbut impair learning [35,86];

skill transfer sessions in daily life: these allow the gener-alization of learned skills [11,35,61]. They may be moreeffective when neurofeedback sessions are conductedwith reference cues to the context of generalization: forexample neurofeedback sessions for ADHD in an envi-ronment reminiscent of a classroom (now possible withportable neurofeedback equipment) [101]; future devel-opments might even combine neurofeedback in virtualenvironments [17].

nstitutional structure

he place of EEG NF within the current therapeuticpproach remains uncertain for the moment. EEG NF doesot properly belong to any single discipline. Neurofeedbacks indeed at the interface between several domains: biomed-cal engineering, neurophysiology and psychophysiology,earning science, cognitive behavioral therapy, cognitiveemediation and health psychology.

EEG NF in neurology and psychiatry is certainly not aherapeutic measure to consider as sole treatment in sub-ects presenting a mental and brain disorder. This is aomplementary non-pharmacological therapy that could besed in conjunction with other treatment modalities. Thenstitutional structure of EEG NF would enable:

promotion of the development of research protocols[36,101,109], which would strengthen the level of ther-apeutic evidence of EEG NF; improve understanding ofneurobiological mechanisms of learning and efficacy ofEEG NF; and identify variables (neurophysiological, cog-nitive, affective and psychosocial) that modulate learningand the efficacy of EEG NF;

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

quality control of the clinical practice of EEG NF [40].This last point requires appropriate training of therapistsusing EEG NF. The Biofeedback Certification InternationalAlliance (BCIA) model is an example that would allow

PRESSJ.-A. Micoulaud-Franchi et al.

controlled and rigorous development of this therapeutictechnique [40].

onclusion

n the fields of neurology and psychiatry, the therapeuticole of EEG NF currently remains controversial. Furthertudies are needed to confirm the effectiveness of EEG NFo treat mental and brain disorders. However, promisingesults have so far been found for therapeutic efficacy inhildren with ADHD and in patients with pharmacoresistantpilepsy. In parallel to the need for controlled, random-zed, blinded studies, recommendations for good practicef this technique and its institutional structure remain to bestablished, as has been done for example for external braintimulation [56].

Despite various obstacles to the implementation of EEGF in clinical practice, the questions raised by this ther-py seem to go beyond the strict field of neurofeedbacknd beyond the controversy between positive and scepticalpinions on the efficacy of EEG NF. Indeed, reflections oneurofeedback are paradigmatic of current thinking in theeurophysiological therapeutic field as applied to psychiatrynd neurology. Use of electrophysiological and neuroimag-ng methods are changing approaches to performing CBT,ognitive remediation [58,82], external brain stimulationreatments [89] and pharmacotherapy [33]. These tech-iques deal with psychophysiological interactions betweenhe subject and new therapeutic neurophysiological toolsvailable to psychiatrist and neurologist; and raise impor-ant questions about how best to evaluate of their efficacy,heir best practice rules and their place within medicalnstitutions, which merit further exploration. Despite itsurrent controversy, EEG NF appears thus to be an exper-mental field that is emblematic of the ongoing evolutionf neurophysiology in the therapeutic field of neurology andsychiatry.

isclosure of interest

he authors declare that they have no competing interest.

eferences

[1] Agnihotri H, Paul M, Singh Sandhu J. Biofeedback approach inthe treatment of generalized anxiety disorder. Iran J Psychi-atry 2007;2:90—5.

[2] American Psychiatryic Association. Diagnostic and statisticalmanual of mental disorder, 5th ed, Text Revision (DSM-5).Washington, DC: American Psychiatryic Association; 2013.

[3] Ang KK, Guan C, Chua KS, Ang BT, Kuah CW, Wang C, et al.A large clinical study on the ability of stroke patients to usean EEG-based motor imagery brain-computer interface. ClinEEG Neurosci 2012;42:253—8.

[4] Arnold LE, Lofthouse N, Hersch S, Pan X, Hurt E, BatesB, et al. EEG-neurofeedback for ADHD: double-blind sham-controlled randomized pilot feasibility trial. J Atten Disord2013;17:410—9.

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

[5] Arns M, Gordon E. Quantitative EEG (QEEG) in psychiatry:diagnostic or prognostic use? Clin Neurophysiol 2014.

[6] Arns M. Historical Archives: the beginning. . .. J Neurother2010;14:291—2.

IN+Model

ARTICLENEUCLI-2498; No. of Pages 11

Electroencephalographic neurofeedback

[7] Arns M, Kenemans JL. Neurofeedback in ADHD and insomnia:vigilance stabilization through sleep spindles and circadiannetworks. Neurosci Biobehav Rev 2014;44:183—94.

[8] Arns M, Strehl U. Evidence for efficacy of neurofeedback inADHD? Am J Psychiatry 2013;170:799—800.

[9] Arns M, de Ridder S, Strehl U, Breteler M, Coenen A. Efficacyof neurofeedback treatment in ADHD: the effects on inatten-tion, impulsivity and hyperactivity: a meta-analysis. Clin EEGNeurosci 2009;40:180—9.

[10] Arns M, Conners CK, Kraemer HC. A decade of EEG Theta/BetaRatio Research in ADHD: a meta-analysis. J Atten Disord2013;17:374—83.

[11] Arns M, Heinrich H, Strehl U. Evaluation of neurofeed-back in ADHD: the long and winding road. Biol Psychol2014;95:108—15.

[12] Bakhshayesh AR, Hansch S, Wyschkon A, Rezai MJ, Esser G.Neurofeedback in ADHD: a single-blind randomized controlledtrial. Eur Child Adolesc Psychiatry 2011;20:481—91.

[13] Barclay-Goddard RE, Stevenson TJ, Poluha W, Thalman L.Mental practice for treating upper extremity deficits in indi-viduals with hemiparesis after stroke. Cochrane Database SystRev 2011:CD005950.

[14] Bauer H, Pllana A. EEG-based local brain activity feedbacktraining-tomographic neurofeedback. Front Hum Neurosci2014;8:1005.

[15] Bink M, van Nieuwenhuizen C, Popma A, Bongers IL, van BoxtelGJ. Neurocognitive effects of neurofeedback in adolescentswith ADHD: a randomized controlled trial. J Clin Psychiatry2014;75:535—42.

[16] Bink M, van Nieuwenhuizen C, Popma A, Bongers IL, van Box-tel GJ. Behavioral effects of neurofeedback in adolescentswith ADHD: a randomized controlled trial. Eur Child AdolescPsychiatry 2015;24:1035—48.

[17] Bioulac S, Lallemand S, Rizzo A, Philip P, Fabrigoule C,Bouvard MP. Impact of time on task on ADHD patient’s per-formances in a virtual classroom. Eur J Paediatr Neurol2012;16:514—21.

[18] Budzynski TH, Budzynski HK, Evans JR, Abarbanel A. Intro-duction to quantitative EEG and neurofeedback. London:Elsevier; 2009.

[19] Cannon RL, Pigott HE, Surmeli T, Simkin DR, Thatcher RW, Vanden Bergh W, et al. The problem of patient heterogeneity andlack of proper training in a study of EEG-neurofeedback inchildren. J Clin Psychiatry 2014;75:289—90.

[20] Carmen J. Passive infrared hemoencephalography: four yearsand 100 migraines. J Neurotherapy 2004;8:23—51.

[21] Choi SW, Chi SE, Chung SY, Kim JW, Ahn CY, Kim HT. Isalpha wave neurofeedback effective with randomized clin-ical trials in depression? A pilot study. Neuropsychobiology2011;63:43—51.

[22] Clemente CD, Sterman MB, Wyrwicka W. Post-reinforcementEEG synchronization during alimentary behavior. Electroen-cephalogr Clin Neurophysiol 1964;16:355—65.

[23] Coben R, Evans JR. Neurofeedback and neuromodulationtechniques and applications. London: Elsevier; 2011.

[24] Coburn KL, Lauterbach EC, Boutros NN, Black KJ, ArciniegasDB, Coffey CE. The value of quantitative electroencephalog-raphy in clinical psychiatry: a report by the Committee onResearch of the American Neuropsychiatric Association. JNeuropsychiatry Clin Neurosci 2006;18:460—500.

[25] Congedo M, Lubar JF, Joffe D. Low-resolution electromagnetictomography neurofeedback. IEEE Trans Neural Syst RehabilEng 2004;12:387—97.

[26] Dagenais E, Leroux-Boudreault A, El-Baalbaki G, Begin J.Doubting the efficacy/effectiveness of electroencephalo-

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

graphic neurofeedback in treating children with attention-deficit/hyperactivity disorder is as yet unjustified. J ClinPsychiatry 2014;75:778—9.

PRESS9

[27] Daly JJ, Wolpaw JR. Brain-computer interfaces in neurologi-cal rehabilitation. Lancet Neurol 2008;7:1032—43.

[28] Davidson RJ. Anterior electrophysiological asymmetries,emotion, and depression: conceptual and methodologicalconundrums. Psychophysiology 1998;35:607—14.

[29] Davidson RJ, Pizzagalli D, Nitschke JB, Putnam K. Depression:perspectives from affective neuroscience. Annu Rev Psychol2002;53:545—74.

[30] Deuschl G, Eisen A. Recommendations for the practice ofclinical neurophysiology: guidelines of the international fed-eration of clinical physiology. Amsterdam: Elsevier; 1999.

[31] Dobkin BH. Brain-computer interface technology as a tool toaugment plasticity and outcomes for neurological rehabilita-tion. J Physiol 2007;579:637—42.

[32] Durup G, Fessard A. L’électroencéphalogramme de l’homme.Observations psycho-physiologiques relatives à l’actiondes stimuli visuels et auditifs. L’année Psychologique1935;36:1—32.

[33] Fink M. Remembering the lost neuroscience of pharmaco-EEG.Acta Psychiatry Scand 2010;121:161—73.

[34] Gevensleben H, Holl B, Albrecht B, Vogel C, Schlamp D, KratzO, et al. Is neurofeedback an efficacious treatment for ADHD?A randomised controlled clinical trial. J Child Psychol Psychi-atry 2009;50:780—9.

[35] Gevensleben H, Rothenberger A, Moll GH, Heinrich H. Neuro-feedback in children with ADHD: validation and challenges.Expert Rev Neurother 2012;12:447—60.

[36] Gevensleben H, Moll GH, Rothenberger A, Heinrich H.Neurofeedback in attention-deficit/hyperactivity disorder —different models, different ways of application. Front HumNeurosci 2014;8:846.

[37] Gruzelier J. Self-regulation of electrocortical activity inschizophrenia and schizotypy: a review. Clin Electroen-cephalogr 2000;31:23—9.

[38] Gruzelier JH. EEG-neurofeedback for optimising perfor-mance. III: a review of methodological and theoreticalconsiderations. Neurosci Biobehav Rev 2013.

[39] Haenschel C, Baldeweg T, Croft RJ, Whittington M, GruzelierJ. Gamma and beta frequency oscillations in response to novelauditory stimuli: a comparison of human electroencephalo-gram (EEG) data with in vitro models. Proc Natl Acad Sci USA2000;97:7645—50.

[40] Hammond D, Bodenhamer-Davis G, CGluck G, Stokes D,Harper S, Trudeau D, et al. Standards of practice forneurofeedback and neurotherapy: a position paper of theInternational Society for Neurofeedback & Research. J Neu-rother 2011;15:54—64.

[41] Hegerl U, Hensch T. The vigilance regulation model ofaffective disorders and ADHD. Neurosci Biobehav Rev2012;44:45—57.

[42] Hoedlmoser K, Pecherstorfer T, Gruber G, Anderer P, Dop-pelmayr M, Klimesch W, et al. Instrumental conditioning ofhuman sensorimotor rhythm (12—15 Hz) and its impact onsleep as well as declarative learning. Sleep 2008;31:1401—8.

[43] Holtmann M, Steiner S, Hohmann S, Poustka L, BanaschewskiT, Bolte S. Neurofeedback in autism spectrum disorders. DevMed Child Neurol 2011;53:986—93.

[44] Hughes JR, John ER. Conventional and quantitative electroen-cephalography in psychiatry. J Neuropsychiatry Clin Neurosci1999;11:190—208.

[45] Jasper H, Shagass C. Conditionning of the occipital alpharythm in man. J Exp Psychol 1941;28:373—88.

[46] John ER, Karmel BZ, Corning WC, Easton P, Brown D, Ahn H,et al. Neurometrics. Science 1977;196:1393—410.

[47] Kamiya J. Biofeedback training in voluntary control of EEG

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

alpha rhythms. Calif Med 1971;115:44.[48] Kim S, Birbaumer N. Real-time functional MRI neurofeedback:

a tool for psychiatry. Curr Opin Psychiatry 2014;27:332—6.

IN+ModelN

1

ARTICLEEUCLI-2498; No. of Pages 11

0

[49] Kluetsch RC, Ros T, Theberge J, Frewen PA, Calhoun VD,Schmahl C, et al. Plastic modulation of PTSD resting-statenetworks and subjective wellbeing by EEG-neurofeedback.Acta Psychiatry Scand 2014;130:123—36.

[50] Kober SE, Witte M, Ninaus M, Neuper C, Wood G. Learning tomodulate one’s own brain activity: the effect of spontaneousmental strategies. Front Hum Neurosci 2013;7:695.

[51] Koprivova J, Congedo M, Raszka M, Prasko J, Brunovsky M,Horacek J. Prediction of treatment response and the effectof independent component neurofeedback in obsessive-compulsive disorder: a randomized, sham-controlled, double-blind study. Neuropsychobiology 2013;67:210—23.

[52] Kotchoubey B, Strehl U, Uhlmann C, Holzapfel S, Konig M,Froscher W, et al. Modification of slow cortical potentials inpatients with refractory epilepsy: a controlled outcome study.Epilepsia 2001;42:406—16.

[53] Lacroix JM. The acquisition of autonomic control throughbiofeedback: the case against an afferent process and a two-process alternative. Psychophysiology 1981;18:573—87.

[54] Lansbergen MM, van Dongen-Boomsma M, Buitelaar JK, Slaats-Willemse D. ADHD and EEG-neurofeedback: a double-blindrandomized placebo-controlled feasibility study. J NeuralTransm 2011;118:275—84.

[55] Lantz DL, Sterman MB. Neuropsychological assessment of sub-jects with uncontrolled epilepsy: effects of EEG feedbacktraining. Epilepsia 1988;29:163—71.

[56] Lefaucheur JP, Andre-Obadia N, Poulet E, Devanne H, HaffenE, Londero A, et al. [French guidelines on the use of repetitivetranscranial magnetic stimulation (rTMS): safety and thera-peutic indications]. Neurophysiol Clin 2011;41:221—95.

[57] Leins U, Goth G, Hinterberger T, Klinger C, Rumpf N, Strehl U.Neurofeedback for children with ADHD: a comparison of SCPand Theta/Beta protocols. Appl Psychophysiol Biofeedback2007;32:73—88.

[58] Linden DE. How psychotherapy changes the brain — thecontribution of functional neuroimaging. Mol Psychiatry2006;11:528—38.

[59] Linden DE. Neurofeedback and networks of depression. Dia-logues Clin Neurosci 2014;16:103—12.

[60] Loomis A, Harvey E, Hobart G. Electrical potentials of thehuman brain. J Exp Psychol 1936;19:249—79.

[61] Mayer K, Wyckoff SN, Strehl U. One size fits all? Slow cor-tical potentials neurofeedback: a review. J Atten Disord2012;17:393—409.

[62] McCarthy-Jones S. Taking back the brain: could neuro-feedback training be effective for relieving distressingauditory verbal hallucinations in patients with schizophrenia?Schizophr Bull 2012.

[63] McKenzie RF, Ehrisman WJ, Montgomery PS, Barnes RH. Thetreatment of headache by means of electroencephalographicbiofeedback. Headache 1974;13:164—72.

[64] Meisel V, Servera M, Garcia-Banda G, Cardo E, Moreno I.Reprint of ‘‘Neurofeedback and standard pharmacologicalintervention in ADHD: a randomized controlled trial with six-month follow-up’’. Biol Psychol 2014;95:116—25.

[65] Micoulaud Franchi J, Geoffroy P, Fond G, Lopez R,Bioulac S, Philip P. EEG-neurofeedback treatments inchildren with ADHD: an updated meta-analysis of ran-domized controlled trials. Front Hum Neurosc 2014,http://dx.doi.org/10.3389/fnhum.2014.00906.

[66] Micoulaud-Franchi JA, Fakra E, Cermolacce M, Vion-DuryJ. [Towards a new approach of neurophysiology in clinicalpsychiatry: functional magnetic resonance imaging neuro-feedback applied to emotional dysfunctions]. Neurophysiol

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

Clin 2012;42:79—94.[67] Micoulaud-Franchi JA, Kotwas I, Lanteaume L, Berthet C,

Bastien M, Vion-Dury J, et al. Skin conductance biofeed-back training in adults with drug-resistant temporal lobe

PRESSJ.-A. Micoulaud-Franchi et al.

epilepsy and stress-triggered seizures: a proof-of-conceptstudy. Epilepsy Behav 2014;41:244—50.

[68] Milstein V. Contingent alpha blocking: conditioning or sensiti-zation? Electroencephalogr Clin Neurophysiol 1965;18:272—7.

[69] Nagai Y, Goldstein LH, Fenwick PB, Trimble MR. Clinicalefficacy of galvanic skin response biofeedback training inreducing seizures in adult epilepsy: a preliminary randomizedcontrolled study. Epilepsy Behav 2004;5:216—23.

[70] Nestoriuc Y, Martin A. Efficacy of biofeedback for migraine: ameta-analysis. Pain 2007;128:111—27.

[71] Nestoriuc Y, Martin A, Rief W, Andrasik F. Biofeedbacktreatment for headache disorders: a comprehensive efficacyreview. Appl Psychophysiol Biofeedback 2008;33:125—40.

[72] Neuper C, Schlogl A, Pfurtscheller G. Enhancement of left-right sensorimotor EEG differences during feedback-regulatedmotor imagery. J Clin Neurophysiol 1999;16:373—82.

[73] Niedermeyer E. Alpha rhythms as physiological and abnormalphenomena. Int J Psychophysiol 1997;26:31—49.

[74] Ogrim G, Hestad KA. Effects of neurofeedback versus stim-ulant medication in attention-deficit/hyperactivity disorder:a randomized pilot study. J Child Adolesc Psychopharmacol2013;23:448—57.

[75] Oken BS, Salinsky MC, Elsas SM. Vigilance, alertness, or sus-tained attention: physiological basis and measurement. ClinNeurophysiol 2006;117:1885—901.

[76] Peeters F, Oehlen M, Ronner J, van Os J, Lousberg R. Neuro-feedback as a treatment for major depressive disorder — apilot study. PLoS One 2014;9:e91837.

[77] Peniston E, Kulkosky P. Alpha-theta brainwave neurofeedbacktherapy for Vietnam veterans with combat-related post-traumatic stress disorder. Med Psychother 1991;4:47—60.

[78] Rayegani SM, Raeissadat SA, Sedighipour L, RezazadehIM, Bahrami MH, Eliaspour D, et al. Effect of neuro-feedback and electromyographic-biofeedback therapy onimproving hand function in stroke patients. Top Stroke Rehabil2014;21:137—51.

[79] Rice KM, Blanchard EB, Purcell M. Biofeedback treatmentsof generalized anxiety disorder: preliminary results. Biofeed-back Self Regul 1993;18:93—105.

[80] Ros TBJB, Lanius RA, Vuilleumier P. Tuning pathological brainoscillations with neurofeedback: a systems neuroscienceframework. Front Hum Neurosci 2014;8:1008.

[81] Schabus M, Heib DP, Lechinger J, Griessenberger H, KlimeschW, Pawlizki A, et al. Enhancing sleep quality and memory ininsomnia using instrumental sensorimotor rhythm condition-ing. Biol Psychol 2014;95:126—34.

[82] Schneider F, Backes V, Mathiak K. Brain imaging: on the waytoward a therapeutic discipline. Eur Arch Psychiatry Clin Neu-rosci 2009;259(Suppl. 2):S143—7.

[83] Schoenberg PL, David AS. Biofeedback for psychiatryic dis-orders: a systematic review. Appl Psychophysiol Biofeedback2014;39:109—35.

[84] Scott WC, Kaiser D, Othmer S, Sideroff SI. Effects of an EEGbiofeedback protocol on a mixed substance abusing popula-tion. Am J Drug Alcohol Abuse 2005;31:455—69.

[85] Servant D. La relaxation : nouvelles approches, nouvellespratiques. Paris; 2009.

[86] Sherlin LH, Arns M, Lubar J, Heinrich H, Kerson C, Strehl U,et al. Neurofeedback and basic lerning therory: implicationsfor research and practice. J Neurother 2011;15:292—304.

[87] Siegle G, Ghinassi F, Thase M. Neurobehavioral therapiesin the 21st century: summary of an emerging field and anextended example of cognitive control training for depres-sion. Cogn Ther Res 2007;31:235—62.

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

[88] Silberstein SD. Practice parameter: evidence-based guide-lines for migraine headache (an evidence-based review):report of the Quality Standards Subcommittee of the Ameri-can Academy of Neurology. Neurology 2000;55:754—62.

IN+Model

[

[

[

[

[

[

[

[

[

[

[

[

ARTICLENEUCLI-2498; No. of Pages 11

Electroencephalographic neurofeedback

[89] Silvanto J, Pascual-Leone A. State-dependency of transcranialmagnetic stimulation. Brain Topogr 2008;21:1—10.

[90] Siniatchkin M, Hierundar A, Kropp P, Kuhnert R, Gerber WD,Stephani U. Self-regulation of slow cortical potentials in chil-dren with migraine: an exploratory study. Appl PsychophysiolBiofeedback 2000;25:13—32.

[91] Skinner BF. Are theories of learning necessary? Psychol Rev1950;57:193—216.

[92] Sonuga-Barke EJ, Brandeis D, Cortese S, Daley D, Ferrin M,Holtmann M, et al. Nonpharmacological interventions forADHD: systematic review and meta-analyses of randomizedcontrolled trials of dietary and psychological treatments. AmJ Psychiatry 2013;170:275—89.

[93] Sonuga-Barke E, Brandeis D, Cortese S, Daley D, Danck-aerts M, Dopfner M, et al. Response to Chronis-Tuscanoet al. and Arns and Strehl. Am J Psychiatry 2013;170:800—2.

[94] Steiner NJ, Sheldrick RC, Gotthelf D, Perrin EC. Computer-based attention training in the schools for children withattention-deficit/hyperactivity disorder: a preliminary trial.Clin Pediatr (Phila) 2011;50:615—22.

[95] Sterman MB. Basic concepts and clinical findings in the treat-ment of seizure disorders with EEG operant conditioning. ClinElectroencephalogr 2000;31:45—55.

[96] Sterman MB, Howe RC, Macdonald LR. Facilitation of spindle-burst sleep by conditioning of electroencephalographicactivity while awake. Science 1970;167:1146—8.

[97] Stewart JL, Bismark AW, Towers DN, Coan JA, Allen JJ. Restingfrontal EEG asymmetry as an endophenotype for depres-sion risk: sex-specific patterns of frontal brain asymmetry.J Abnorm Psychol 2010;119:502—12.

[98] Stoeckel LE, Garrison KA, Ghosh S, Wighton P, Hanlon CA,Gilman JM, et al. Optimizing real time fMRI neurofeedbackfor therapeutic discovery and development. Neuroimage Clin2014;5:245—55.

[99] Stokes DA, Lappin MS. Neurofeedback and biofeedback with

Please cite this article in press as: Micoulaud-Franchi J-A,

evidence in mental and brain disorders and suggestions for gNeurophysiology (2015), http://dx.doi.org/10.1016/j.neucli.20

37 migraineurs: a clinical outcome study. Behav Brain Funct2010;6:9.

[100] Strehl U. Slow cortical potentials neurofeedback. J Neurother2009;13:117—26.

PRESS11

101] Strehl U. What learning theories can teach us in designingneurofeedback treatments. Front Hum Neurosci 2014;8:894.

102] Strijkstra AM, Beersma DG, Drayer B, Halbesma N, Daan S.Subjective sleepiness correlates negatively with global alpha(8—12 Hz) and positively with central frontal theta (4—8 Hz)frequencies in the human resting awake electroencephalo-gram. Neurosci Lett 2003;340:17—20.

103] Sulzer J, Haller S, Scharnowski F, Weiskopf N, Birbaumer N,Blefari ML, et al. Real-time fMRI neurofeedback: progress andchallenges. Neuroimage 2013;76:386—99.

104] Surmeli T, Ertem A, Eralp E, Kos IH. Schizophrenia and theefficacy of qEEG-guided neurofeedback treatment: a clinicalcase series. Clin EEG Neurosci 2012;43:133—44.

105] Tan G, Thornby J, Hammond DC, Strehl U, Canady B, Arne-mann K, et al. Meta-analysis of EEG biofeedback in treatingepilepsy. Clin EEG Neurosci 2009;40:173—9.

106] Thatcher LG, Barlow JF. Thirteen-year-old girl with one monthhistory of back pain and limp. S D J Med 1976;29:31—5.

107] van Dongen-Boomsma M. Dr van Dongen-Boomsma replies. JClin Psychiatry 2014;75:779.

108] van Dongen-Boomsma M, Vollebregt MA, Slaats-Willemse D,Buitelaar JK. Dr van Dongen-Boomsma and colleagues reply.J Clin Psychiatry 2014;75:290.

109] Vollebregt MA, van Dongen-Boomsma M, Slaats-Willemse D,Buitelaar JK. What future research should bring to help resolv-ing the debate about the efficacy of EEG-neurofeedback inchildren with ADHD. Front Hum Neurosci 2014;8:321.

110] Witte M, Kober SE, Ninaus M, Neuper C, Wood G. Controlbeliefs can predict the ability to up-regulate sensorimotorrhythm during neurofeedback training. Front Hum Neurosci2013;7:478.

111] Wolpaw JR, Birbaumer N, McFarland DJ, Pfurtscheller G,Vaughan TM. Brain-computer interfaces for communicationand control. Clin Neurophysiol 2002;113:767—91.

112] Zuberer A, Drandeis D, Drechsler R. Are treatment effectsof neurofeedback training in children with ADHD related to

et al. Electroencephalographic neurofeedback: Level ofood clinical practice. Neurophysiologie Clinique/Clinical15.10.077

the successful regulation of brain activity? A review on thelearning of regulation of brain activity and a contributionto the discussion on specificity. Front Hum Neurosc 2015,http://dx.doi.org/10.3389/fnhum.2015.00135.